i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 5 2 ( 2 0 2 4 ) 1 1 1 3 e1 1 2 3

Available online at www.sciencedirect.com

ScienceDirect

journal homepage: www.elsevier.com/locate/he

Hydrogenation of liquid organic hydrogen carriers:

Process scale-up, economic analysis and
optimization

n Garcı́a, Pablo Marı́n, Salvador Ordo

Adria
n~ ez*
Catalysis, Reactors and Control Research Group (CRC), Dept. of Chemical and Environmental Engineering, University

n Claverı́a 8, Oviedo 33006, Spain
of Oviedo, Julia

highlights

Process evaluated for 3,5-dibenzyl toluene, 2-methyl indole, N-ethylcarbazole LOHCs.

Process scale-up and optimization for LOHC hydrogenation: 190  C and 6 MPa.

Economic evaluation suggests that 3,5-dibenzyl toluene is the best LOHC.

Cost of LOHC make-up is the most important for the H2 storage cost.

article info abstract

Article history: Hydrogen has been proposed as energy vector because it can be produced from renewable
Received 25 July 2022 energy sources and its use produces no carbon dioxide. Moreover, many industrial pro-
Received in revised form cesses (e.g., steel industry, ammonia, and fertilizer industry, etc.) are also adopting
8 May 2023 renewable-produced hydrogen as raw material, replacing traditional fossil sources. In this
Accepted 24 June 2023 context, the storage and transportation of hydrogen using liquid organic hydrogen carriers
Available online 10 July 2023 (LOHC) is interesting to deal with the discontinuous production of renewable energies.
The work addressed the study of the hydrogenation process for three compounds, 3,5-
Keywords: dibenzyl toluene, 2-methyl indole and N-ethylcarbazole, on Ni and Ru-based catalysts. The
Hydrogen reactor consisted of a three-phase slurry stirred tank reactor, which has been scale-up
Energy storage together with the rest of the process equipment units using Aspen Hysys simulation
Hydrogenation software. The hydrogen storage cost has been determined applying an economic evalua-
Economic assessment tion to the overall process. The most important costs of the process are the reactor cost and
Optimization the cost of raw materials (the liquid organic hydrogen carrier). The main design variables,
plant capacity, reactor temperature and pressure, were optimized to reduce the storage
cost. The liquid organic hydrogen carrier with the minimum storage cost at the optimum
operating conditions is 3,5-dibenzyl toluene with a reactor configuration of two parallel
tanks.
© 2023 The Author(s). Published by Elsevier Ltd on behalf of Hydrogen Energy Publications
LLC. This is an open access article under the CC BY-NC-ND license (http://
creativecommons.org/licenses/by-nc-nd/4.0/).

* Corresponding author.

n
E-mail address: sordonez@uniovi.es (S. Ordo ~ ez).
https://doi.org/10.1016/j.ijhydene.2023.06.273
0360-3199/© 2023 The Author(s). Published by Elsevier Ltd on behalf of Hydrogen Energy Publications LLC. This is an open access article under the CC BY-NC-
ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
1114 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 5 2 ( 2 0 2 4 ) 1 1 1 3 e1 1 2 3
Introduction
After decades of increasing CO2 emissions to the atmosphere,
world climate is being affected at large scale. From the Artic to
the Equatorial regions, temperature and climate instability is
increasing. The use of fossil fuels for energy production and
industrial purposes is the main source of anthropogenic CO2
emissions. Furthermore, the availability of fossil resources
(e.g., petroleum and natural gas) is concentrated in few
countries. For this reason, the actual energetic model creates
strong dependencies and strains in the global geopolitics,
supports dictatorships that deny human rights, and use the
fuel supply as a weapon. In this context, the European Union
is determined to decarbonize the economy and abandon the
use of fossil fuels, as soon as possible [1].
Several measurements have been proposed and imple-
mented to reduce CO2 emissions, such as, the increase of ef-
ficiency and save of energy, the promotion of renewable
energies and biofuels, etc. [2,3]. One important problem of
most of the renewable energies in expansion, like wind and
solar, is the high variability in their production. For this reason,
different methodologies are being developed to store energy,
when there is an excess of generation, and to use this excess
when there is a deficit of generation. The electrical energy
must be transformed into other type of energy for the storage,
for instance, mechanical energy in damps and pressurized
vessels or chemical energy in electrical batteries. In these
systems, during the discharge stage the energy is recovered as
electrical energy, which can be introduced in the network.
However, these forms of energy storage might not be suitable
for the industry, which typically consumes considerable
amounts of energy. In many cases, like in chemical, food and
process industries, the energy is consumed as thermal energy,
so the use of hydrogen as energy carrier is a better choice [4e7].
Hydrogen can be produced by electrolysis during periods of
generation surplus from renewable sources. Hydrogen is a
viable candidate as energy carrier: it is a concentrated energy
source, it does not produce pollutants during utilization and it
can be used as raw material, replacing other fossil sources. In
the chemical industry, hydrogen is used as reducing agent for
the manufacture of many chemicals, such as, ammonia (used
to produce fertilizers), methanol (used as solvent, production
of special chemicals, polymers, and resins), hydrogenation of
organic compounds, etc. In the iron and steel industry, the use
of hydrogen in the direct reduction of iron process is called to
replace the traditional blast furnace process based on the use
of coke [8,9].
The storage of hydrogen is particularly important to couple
a discontinuous renewable production with a continuous in-
dustrial use. There have been proposed different alternatives
of storage, as summarized below [10e12]. The physical
methods consists on the storage of hydrogen in tanks, as a
compressed gas (50e70 MPa) or liquefied (253  C) [13]. The
main drawbacks of these methods are safety and the low
energy density of the storage.
The adsorption methods are based on the use of solids,
typically porous materials with high surface area, capable of
adsorbing hydrogen of their surface. Materials, like active
carbon, carbon nanotubes or zeolites, have been investigated;
however, the conditions required to storage (200  C and
5.5 MPa) limit their practical applications [14e16].
The reaction methods use hydrogen as reactant to
generate chemical compounds suitable for the storage. The
reactions involved must be reversible to recover the hydrogen
molecule from the compound when needed. The use of metal
hydrides (like LiBH4) [10,17], ammonia, methanol, formic acid
and unsaturated liquid organic compounds has been investi-
gated [11,18,19]. Ammonia is very interesting due to its high
hydrogen density and the direct applications in industry
[7,18,20]. Methanol and formic acid have moderate storage
capacity and they can be generated directly by electrolysis
[21]; however, at the point of use, where hydrogen is released,
CO2 is emitted [22,23].
The unsaturated liquid organic hydrogen carriers (LOHC)
are organic compounds capable of reacting with hydrogen to
generate a saturated form. The hydrogenation reaction is an
exothermic reaction, requiring high pressure, moderate tem-
perature, and the use of a catalyst. The reverse dehydroge-
nation is an endothermic catalysed reaction, favoured at low
pressure (typically atmospheric) and high temperature [24,25].
The main advantages of unsaturated LOHC are their handling
easiness, safety, and the possibility of using the actual infra-
structure for petroleum oil and derived fuels. This is very
interesting to reduce the investment and help to redeem the
systems and installations in use with fossil fuels [26,27].
The compounds considered as LOHC must fulfil some re-
quirements, such as, high capacity (minimum 5 wt %), avail-
ability (easy to produce and inexpensive compounds), low
toxicity, ease of handling (high boiling and low freezing points,
stability), and low dehydrogenation temperature (too high
temperature would increase considerably the operational costs)
[28]. Most commonly proposed LOHC are aromatic compounds
(benzene, toluene, benzyl toluene, dibenzyl toluene, naphtha-
lene or biphenyl) [24,29e32] and heterocyclic compounds
(quinoline, 2-(N-methylbenzyl)pyridine, N-ethylcarbazole or 2-
methyl indole) [33e37]. The aromatic compounds have a
higher hydrogen storage capacity (>6 wt %), but require a higher
temperature for dehydrogenation (more than 270  C, against
150-190  C of the heterocycle compounds) [26,38,39].
Although there are many published papers about the
screening of chemicals and catalysts for being used for this
purpose, there is a general lack of information about the
designing of these units. It should be considered that, although
the involved reactions are known from other applications, re-
action conditions are quite different in the LOHC application.
Aspect as the absence of solvents in the hydrogenation steps
completely modify the design of these reactors. Thus, the
scope of this work is to perform a detailed design and opti-
mization of the hydrogenation step (LOHC loading) using three
different commonly used LOHCs (3,5-dibenzyl toluene, 2-
methyl indole and N-ethylcarbazole). These products are
commonly proposed in the literature as LOHCs because of their
physics-chemical properties, summarized in Table 1.
The first compound, 3,5-dibenzyl toluene, is an aromatic
hydrocarbon with industrial applications as thermal fluid.
This compound can be hydrogenated into perhydro-dibenzyl
toluene at 140  C and up to 70 MPa using supported precious
metal catalysts (e.g., Ru/Al2O3 or Pt/Al2O3) without formation
of secondary or degradation products [32,40e42].
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 5 2 ( 2 0 2 4 ) 1 1 1 3 e1 1 2 3 1115

Table 1 e Properties of the LOHC studied in the present work.

Name 3,5-dibenzyl toluene 2-methyl indole N-ethylcarbazole
Formula

Freezing 39  C 58  C 70  C
Boiling 390  C 273  C 295  C
Stoichiometry coef. 9 mol H2/mol LOHC 4 mol H2/mol LOHC 6 mol H2/mol LOHC
Reaction enthalpy 530 kJ/mol 206 kJ/mol 342 kJ/mol
Weight capacity 6.2 wt% 5.8 wt % 5.6 wt%
Volumetric capacity 762 L H2/L 705 L H2/L 736 L H2/L
H2 solubility at 1 MPa, 150  C 17.6 mol/m3 16.3 mol/m3 17 mol/m3

The other two compounds, 2-methyl indole and N-ethyl-
carbazole, are both heterocycle compound with equivalent
properties. Like for 3,5-dibenzyl toluene, the hydrogenation
takes place using supported precious metal catalysts at
similar conditions [36,37,43e45]. These compounds are also
stable and non-toxic, but the hydrogenated forms have high
freezing points, which can complicate the storage. The main
advantage of these compounds is their lower dehydrogena-
tion temperature [34].
The proposed reactor consisted of a continuous three-phase
stirred tank slurry reactor. Hydrogen is fed in excess, so
unreacted hydrogen is separated and recycled to the reactor.
This reactor type is more advantageous for heat and mass
transfer. The hydrogenation reactions are very exothermic (see
reaction enthalpies reported in Table 1) and, given that no
solvents are used, temperature must be carefully controlled.
Stirred tanks equipped with a jacked for refrigeration are sim-
ple and efficient. In stirred tank reactors, the catalyst is in the
form of a slurry. To facilitate the suspension, the catalyst par-
ticles must be of small size, which means that mass transfer to
the particles and in the inside is fast. The use of a stirred also
helps to transfer and dissolve hydrogen in the liquid phase. For
these reasons, the use of stirred tank reactors is a good option
to carry out the hydrogenation process at the industrial scale.
The reactor scale-up has been based on a kinetic model of
the respective hydrogenation catalytic reactions; the influ-
ence of mass transfer on reaction rate has been estimated and
discussed. In the economic evaluation section, the capital
investment and operating costs of the process are estimated
and used to determine the unitary cost of hydrogen storage.
Finally, the main operating and design variables of the process
are optimized by means of a sensitivity analysis with the aim
of decreasing the unitary cost of hydrogen storage.
Material and methods
Kinetic model
The catalytic hydrogenation of the LOHC is a three-phase re-
action involving gaseous hydrogen, liquid LOHC and the solid
catalyst. This reaction is accomplished according to three
steps in series: (1) hydrogen transfer and dissolution in the
liquid phase, (2) hydrogen transfer to the catalyst surface and
(3) reaction on the catalyst surface. Using the film theory to
describe the mass transfer steps and assuming a reaction ki-
netics of first order with respect to each reactant (LOHC and
H2), the following equations can be formulated:

 
rH2 ¼ akL pH2 HH2  CH2 liq (1)


rH2 ¼ fs aP kC CH2 liq  CH2 sup (2)
rH2
rLOHC ¼ ¼ fs kv CLOHC CH2 sup (3)
nH2
Where rH2 and rLOHC are, respectively, the hydrogen and
LOHC, reaction rates per unit volume of reactor, nH2 is the
stoichiometric coefficient of H2 in the hydrogenation reaction
(see Table 1), akL is the gas-to-liquid mass transfer coefficient,
aP kC is the liquid-to-solid mass transfer coefficient, kv is the
kinetic constant per unit volume of catalyst (i.e., kv ¼ kw rcat ),
pH2 is the partial pressure of hydrogen in the gas phase, HH2 is
the Henry's law constant for hydrogen, CH2 liq and CH2 sup are,
respectively, the molar hydrogen concentrations in the bulk
liquid and the surface of the catalyst, CLOHC is the molar con-
centration of LOHC, and fs is the volumetric fraction of catalyst
in the reactor.
Equations (1)e(3) have been solved for CH2 liq and CH2 sup , to
obtain an expression of the reaction rate as a function of
hydrogen partial pressure, the transport coefficients and the
kinetic constant. The reaction rate of the LOHC is given by the
following expression, where each term of the denominator
corresponds to a resistance (respectively, mass transfer to the
liquid phase, mass transfer to the catalyst surface and
reaction):
 . 
pH2 HH2
rH2
rLOHC ¼ ¼ nH2 n (4)
nH2 akL
þ fs aHP2kC þ fs kv C1LOHC
For the case of fast mass transfer steps with respect to the
reaction, eq. (4) can be simplified eliminating the first two
terms of the denominator:
1116 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 5 2 ( 2 0 2 4 ) 1 1 1 3 e1 1 2 3

.
rLOHC ¼ fs kv CLOHC pH2 HH2 (5) Results and discussion
The reactions considered in the present work are the hy-
Assessment of mass transfer limitations
drogenation of 3,5-dibenzyl toluene on Ni Raney, and 2-
methyl indole and N-ethylcarbazole on Ru/Al2O3. The kinetic
The reactor selected to carry out the three-phase hydrogena-
constants have been calculated using reaction experiments
tion reaction consists of a slurry stirred tank. This reactor is
from the bibliography, which were carried out in the absence
equipped with a mechanical stirrer to maintain the solid
of mass transfer resistances [41,44,45]. Arrhenius-type
catalyst in suspension and a diffuser to disperse hydrogen gas
dependence with temperature of the kinetic constant is
in the liquid LOHC. In order to help the suspension of the
assumed, kv ¼ kv0 eEa =RT , where kv0 is the pre-exponential
catalyst, a particle size of 0.1 mm and a fraction catalyst of
factor, Ea is the activation energy, R is the ideal gas constant
0.5 vol % have been selected [47]. As shown in eqs. (4) and (5),
and T is temperature. Table 2 shows the fitted parameters of
higher catalyst fractions (fw ) result in higher reaction rates but
the kinetic model, eq. (5).
hindering the catalyst-liquid mixing. In these multi-phase
systems, good mixing inside the reactor is ensured using, at
Process simulation
least, a stirring speed of 90 rpm and a specific power of 1.6 kW/
m3 [47].
The study of the hydrogenation process has been carried out
The mass transfer coefficients of eq. (4) have been deter-
by simulating with the software Aspen Hysys. Peng-Robinson
mined using correlations for stirred tanks [48,49]. These cor-
thermodynamic model was selected, and the properties of the
relations have been evaluated using the above parameters
hydrogenated compounds were introduced in the model.
and the physical properties, estimated at a temperature of
Fig. 1 shows the flowsheet of the hydrogenation process.
190  C. The results of the calculations are summarized in
The hydrogenation reactor is a three-phase slurry reactor,
Table 3. As shown, the estimated mass transfer resistances
equipped with a mechanical stirrer to maintain the solid
are much lower than the reaction one. This means that the
catalyst in suspension and a diffuser to disperse hydrogen gas
reaction step is the slowest one and controls the overall re-
in the liquid LOHC. This reactor has been modelled using the
action rate. The main operating and design variables affecting
continuous stirred tank reactor model, which assumes perfect
the mass transfer resistances are stirring speed (90 rpm) and
mixture inside the reactor. The kinetic model of section 2.1
power (1.6 kW/m3), catalyst particle size (0.1 mm) and catalyst
has been used to predict reaction rates. Two cases have been
fraction (0.5 vol %). The values considered for these variables
considered, a reactor made of one tank and two tanks in se-
were chosen to maintain good catalyst suspension and mix-
ries; the latter one corresponds to the flowsheet shown in
ing, and, at the same time, they were enough to make mass
Fig. 1.
transfer resistances negligible.
Hydrogen fed to the process is produced in an electrolytic
cell using renewable electricity and water. Typical electrolytic
Process scale-up and simulation
cells operate at 6 MPa and 80  C [46], so these are the conditions
chosen for the hydrogen feed stream. The reactor operates
The hydrogenation process has been simulated and scaled up
using hydrogen in excess to shift the reaction towards products.
considering the kinetic model presented in section 2.2. The
However, given its low solubility, not all the hydrogen is dis-
nominal capacity supposed for the process is 6 MW of
solved in the liquid. As shown in Fig. 1, the hydrogen streams
hydrogen power, which corresponds to a H2 consumption of
leaving the reactors without dissolving and reacting are recy-
1578 t/year. According to the reaction kinetics, a reactor
cled and mixed with the fresh feed. This loop of hydrogen is
temperature of 170  C and pressure of 6 MPa are appropriate to
important to maintain the required excess of hydrogen and a
operate the reactor, so these values were fixed as a first esti-
good dispersion of the gas inside the liquid phase.
mation. These variables will be optimized in the sensitivity
The LOHC is introduced in the tanks using pumps. At the
analysis of section 3.4.
end of the process, the hydrogenated form of the LOHC is
Considering these specifications, the reactor volume
cooled in a heat exchanger and expanded to atmospheric
required to achieve 99% conversion of the LOHC was calcu-
pressure. This causes that a bit of dissolved unreacted
lated. A height of liquid inside the tank equals to the diameter
hydrogen desorbs from the liquid phase. This gas can be
is assumed to size the tank. For the case of 3,5-dibenzyl
vented or, otherwise, collected, recompressed, and recycled to
toluene and 2-methyl indole, it is considered the possibility
the reactor.
of using either a reactor made of one tank or, otherwise, two
tanks in series. The latter option is interesting to reduce the
Table 2 e Hydrogenation kinetic model of the LOHC used total reactor volume for the same final conversion, because a
in the present work [41,44,45]. higher LOHC concentration in the first tank increases the re-
action rate. For N-ethylcarbazole, the reaction rate is lower
3,5-dibenzyl 2-methyl N-
toluene indole ethylcarbazole and the option of using only one tank is discharged, because of
the large volume required.
Catalyst Ni-Raney Ru/Al2O3 Ru/Al2O3
The hydrogenation reactions are exothermic (the en-
ln kv0 (m3/mol s) 12.1 18.2 22.5
Ea (kJ/mol) 48.4 73.8 93.9 thalpies of the reaction are displayed in Table 1), so the re-
Temperature (ºC) 150e200 120e170 120e170 actors require refrigeration to maintain the desired reaction
Source data [41] [44] [45] temperature. The refrigeration of the reactors is done using
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 5 2 ( 2 0 2 4 ) 1 1 1 3 e1 1 2 3 1117

Fig. 1 e Flowsheet of the hydrogenation process.

Economic evaluation
Table 3 e Assessment of mass transfer limitations in the
slurry reactors at 190  C.
The aim of this section is to carry out an estimation of the
3,5- 2- N- costs of the process, in order to calculate the cost of H2 stor-
dibenzyl methyl ethylcarbazole age. The cost of the process is divided into capital investment
toluene indole
and operating costs.
Kinetic constant, kv 0.6 0.2 0.01 The capital investment has been estimated as a function of
(m3/mol s)
the cost of the main pieces of equipment (reactors, separator,
Gas-to-liquid mass transfer 0.096 0.12 0.10
compressor, etc.). The purchase cost of the equipment units
coefficient, akL (s1) [48]
Liquid-to-solid catalyst 215 255 232 has been calculated using correlations from the bibliography
mass transfer coefficient, [50], which depends of the corresponding design variables
aP kC (s1) [49] displayed in Table 4. The purchase costs have been updated to
Resistances (s) 2021 using the Chemical Engineering Process Cost Index
Gas-to-liquid, 1= akL 10.4 8.3 10 (CEPCI). The delivery cost of the equipment units is estimated
Liquid-to-solid, 1= fs aP kC 0.93 0.78 0.86
as 10% of the purchase cost. The capital investment includes
Reaction, 1= fs kv CLOHC 7.1 $ 104 21.3 $ 104 426 $ 104
the fixed capital (cost of equipment, installation, piping,
instrumentation, electrical, building, services, engineering,
cooling water that flows through the reactor jacket, given that contingencies, etc.) and working capital (estimated as 15% of
the use of coils is not recommended for slurry reactors. The the total investment). The total capital investment is calcu-
area needed for the reactor jacket is calculated using an lated as 6.05 times the total cost of the delivered equipment
overall heat transfer coefficient of 0.5 kW/m2 K. In case this units. Further details can be found in the bibliography [51].
area is larger than the available in the tank, its geometry is The results of these estimations are summarized in Table 5.
adjusted to fulfil with the cooling requirements. Table 4 shows The LOHC process requiring the lowest capital investment
the results of the scale-up for the main units of equipment. is the one based on 3,5-dibenzyl toluene and a reactor of one
The power consumed by the impeller of the reactor, the tank. The options based on a two-tank reactor require, in
compressor and pumps has also been calculated. This is general, a higher capital investment. The piece of equipment
needed to estimate the consumption of electricity and the with the higher contribution to the total cost is the reactor
operating costs. (40e50%), so the cost of an additional tank has an important
influence. The process based on N-ethyl carbazole requires
the higher capital investment.
Table 4 e Scale-up of the main units of equipment.
The total investment has been annualized considering a
3,5- 2-methyl N-ethyl 10-year period and an interest rate of 15%; the corresponding
dibenzyl indole carbazole annual capital charge ratio (ACCR) is 0.199 [50].
toluene
The operating costs are divided into fixed cost (insurances
1 2 1 2 2 and taxes) and variable costs (raw materials, labour, utilities,
tank tanks tank tanks tanks etc.). The estimates of these costs according to the method-
H2 feed (kg/h) 185 186 180 190 182 ology proposed in the bibliography [51] is summarized in
LOHC flow rate (m3/h) 3.35 3.35 3.28 3.42 3.35 Table 6. The main costs correspond to raw materials, utilities
Reactor size (m3, each 8 8 8 6 6
(cooling water, 2.17 V/1000 kg) and electricity (0.0645 V/kWh).
tank)
Regarding the raw materials, it is considered that most of
Compressor (kW) 21.6 23.7 23 20 115
Pump (kW) 7.5 7.5 7.4 7.7 7.6 the LOHC can be re-used after H2 is released in the dehydro-
Heat exchanger area 12 13 13 13 12 genation process. For estimating purposes, 90% of the LOHC
(m2) feed flow rate would correspond to re-used compound and the
Separator (m3) 0.31 0.31 2.6 0.43 1.52 remaining 10% to make-up. The make-up would compensate
1118 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 5 2 ( 2 0 2 4 ) 1 1 1 3 e1 1 2 3

Table 5 e Summary of the capital investment calculations (currency in Euro).

INVESTMENT 3,5-dibenzyl toluene 2-methyl indole N-ethyl carbazole
1 tank 2 tanks 1 tank 2 tanks 2 tanks
Reactors 112 500 225 000 113 000 191 000 191 000
Agitation 18 000 53 600 53 600 45 100 90 100
Compressors 39 000 56 000 54 700 51 000 126 000
Pumps 4900 9900 4800 9700 9600
Heat exchangers 29 700 51 200 29 800 51 300 30 300
Separator 36 700 36 700 37 600 36 700 37 000
Total equipment cost 240 800 432 400 293 500 384 800 484 000
TOTAL CAPITAL INVESTMENT 1 623 300 2910 000 1 970 000 2590 000 3 260 000
ANNUALIZED INVESTMENT 323 000 579 000 392 000 516 000 648 000

Table 6 e Summary of the operating costs calculations (currency in Euro).

OPERATING COSTS 3,5-dibenzyl toluene 2-methyl indole N-ethyl carbazole
1 tank 2 tanks 1 tank 2 tanks 2 tanks
Fixed costs
Insurances 27 600 49 500 33 500 44 000 55 400
Taxes 13 800 24 700 16 800 22 000 27 700
Variable costs
Raw materials 2 217 000 2220 000 2 250 000 2350 000 4 290 000
Labour 771 000 771 000 771 000 771 000 1 030 000
Supervision 116 000 115 000 115 000 116 000 154 000
Utilities 1378 000 1 390 000 1030 000 1 080 000 1390 000
Electricity 40 100 64 500 43 700 47 800 160 000
Maintenance 96 600 173 000 117 000 154 000 194 000
Supplies 14 500 26 000 17 600 23 000 29 000
Laboratory 116 000 116 000 116 000 116 000 154 000
Cannon 88 700 88 700 89 900 94 000 171 000
Catalyst 2600 2600 400 000 200 000 300 000
TOTAL OPERATING COST 4 881 900 5041 000 5 000 500 5017 800 7 955 100

losses of LOHC in the transportation, storage, or hydrogena-
tion/dehydrogenation processes. Therefore, only the cost of
that 10% of LOHC is accounted for in the operating costs of
Table 6. The prices of the LOHC are 0.90 V/kg for 3,5-dibenzyl
toluene, 0.86 V/kg for 2-methyl indole, and 1.64 V/kg for N-
ethyl carbazole. It should be noted that as the storage tech-
nology is spread and optimized, the LOHC recovery may in-
crease and the amount of make-up considerably reduced,
improving the economy of the process.
The operating costs are quite similar for 3,5-dibenzyl
toluene and 2-methyl indole for the different process config-
urations with one and two tanks. However, they are consid-
erably higher for N-ethyl carbazole, which is due to the higher
price of this LOHC.
Table 7 shows a summary of the total annual costs,
including the annualized investment and the operating costs,
for the different alternative LOHC. The lowest total annual
cost corresponds to 3,5-dibenzyl toluene and the case of using
a reactor of one tank. The cost of H2 storage has also been
calculated using the actual plant capacity and H2 consump-
tion. The nominal design H2 plant capacity was fixed to 6 MW.
However, after the scale-up, the actual plant capacity can be
slightly higher (deviations are within 5%). This is mainly due
to reactor over-sizing to ensure an adequate heat removal to
the tank jacket.
For 3,5-dibenzyl toluene, the cost of H2 storage is 3.22 V/kg
for the one-tank configuration. Nevertheless, the costs for 3,5-
dibenzyl toluene and 2-methyl indole are remarkably close, in
the range 3.22e3.45 V/kg. The cost of N-ethyl carbazole is
clearly higher 5.38 V/kg.
Sensitivity analysis
The aim of the sensitivity analysis is the study of the main
operation and design variables on the production cost of
hydrogen. The considered variables are plant capacity
(measured in terms of hydrogen power), reactor temperature
and pressure. These variables have been varied one at a time
with respect to the base case of sections 3.2 and 3.3 (respec-
tively, 6 MW of hydrogen power, 170  C and 6 MPa). For each
combination of variables, the main pieces of equipment of the
plant (i.e., reactor, vessels, heat exchangers and pumping)
have been sized, the capital and operational costs determined,
and the production cost of hydrogen calculated, as explained
in sections 3.2 and 3.3.
Fig. 2 (left) shows the results for the process based on 3,5-
dibenzyl toluene as LOHC. The configuration of a reactor
made of one and two tanks has been compared in terms of
the production cost of hydrogen. The plant capacity has the
highest influence on the production cost of hydrogen, which
decreases sharply in the range 2e10 MW; above 10 MW, with
H2 costs around 3 V/kg, the influence is less marked. The
curves for one and two tanks are close, being the configura-
tion made of one tank more favourable. This curve is
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 5 2 ( 2 0 2 4 ) 1 1 1 3 e1 1 2 3 1119

Table 7 e Summary of process costs and calculation of H2 storage cost (currency in Euro).
3,5-dibenzyl toluene 2-methyl indole N-ethyl carbazole
1 tank 2 tanks 1 tank 2 tanks 2 tanks
ANNUALIZED INVESTMENT 323 000 579 000 392 000 516 000 648 000
OPERATING COST 4 881 900 5041 000 5 000 500 5017 800 7 955 100
TOTAL ANNUAL COST 5 204 900 5620 000 5 392 500 5533 800 8 603 100
H2 plant (MW) 6.2 6.2 6.0 6.3 6.1
H2 consumption (t H2/year) 1630 1630 1580 1660 1600
Storage cost (V/kg H2) 3.22 3.45 3.41 3.32 5.38

Fig. 2 e Sensitivity analysis of the main design variables on hydrogen cost. Left graphs: 3,5-dibenzyl toluene with a reactor
of one ( ) and two ( ) tanks in series. Right graphs: 2-methyl indole with a reactor of one ( ) and two ( ) tanks in series.
1120 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 5 2 ( 2 0 2 4 ) 1 1 1 3 e1 1 2 3

particularly useful to select the plant capacity for a given

application.
The reactor temperature has almost no influence on the
production cost of hydrogen, being the configuration with one
tank the best option. On increasing reaction temperature, the
reaction rate increases (according to Arrhenius equation and
the kinetic parameters of Table 2), which may lead to a lower
reactor size and cost. However, the LOHC 3,5-dibenzyl toluene
has a remarkably high reaction enthalpy (530 kJ/mol, Table
1) and the reactor must be designed to extract all this heat
and maintain the required temperature. For this reason, the
reactor volume cannot be reduced and there is no benefit
when increasing temperature.
Finally, the reactor pressure has no influence on H2 cost
up to 6 MPa and, above this value, the cost increases linearly.
This behaviour is explained because hydrogen is supplied
from an electrolytic cell at 6 MPa. Thus, above 6 MPa, a
compressor is required to raise pressure and, hence, the
costs increase.
The sensitivity analysis for 2-methyl indole is shown in
Fig. 2 (right). The influence of plant capacity on hydrogen cost
has a similar curve to that of 3,5-dibenzyl toluene. At low plant
capacity values, the cost decreases sharply and, for high
values, it decreases asymptotically to a value around 3 V/kg.
For this LOHC, there is no difference between one and two
tanks since their curves overlap.
In this case, the reactor temperature has a positive in-
fluence on the cost of H2, particularly for the configuration
with one tank. This LOHC has a lower reaction enthalpy,
and the reactor design is ruled by the reaction kinetic, rather
than the heat exchange to the tank jacket. Thus, on
increasing the reaction temperature, reaction rate increases
and, hence, the reactor volume and cost decreases, leading
to a lower H2 cost. For a reactor configuration of two tanks,
this behaviour is not observed, because the reduction in the
total reactor volume obtained with two tanks cannot be
implemented due to the requirements of the reaction heat
removal. Regarding the reactor pressure, the behaviour is
quite similar to 3,5-dibenzyl toluene with an increase of H2
cost above 6 MPa.
The sensitivity analysis of the last LOHC, N-ethylcarbazole,
for the case of a reactor configuration of two tanks is shown in
Fig. 3. The curves of the other LOHCs for the same reactor
configuration have also been included. It can be observed that
the overall behaviour of N-ethylcarbazole can be extrapolated
from that of the other LOHC. However, the H2 cost scale range
is much higher for this LOHC. The reactor temperature has a
small influence on the cost, which decrease on increasing
temperature. This can be explained by the lower reaction rate
of this LOHC, which means that the reactor sizing is ruled by Fig. 3 e Sensitivity analysis of the main design variables on
the reaction kinetics. The trends of plant capacity and reactor hydrogen cost for a reactor made of two tanks in series.
pressure are similar to the other LOHC. Comparison of LOHC: ( ) 3,5-dibenzyl toluene, ( ) 2-
Overall, it can be said that 2-methyl indole is the LOHC methyl indole and ( ) N-ethylcarbazole.
providing the lowest H2 storage cost; however, 3,5-dibenzyl
toluene performs similar with a slightly higher cost.
The optimum design variables are plant capacity of 18 MW, reduction on the cost. However, larger plants required higher
reactor temperature of 190  C and pressure of 6 MPa. As shown capital investment and have a reduced operability, e.g., to
in Fig. 3, the plant capacity has an asymptotic behaviour, adapt to market variations. For this reason, 18 MW is selected
which means that an increase above 18 MW produces little as an adequate optimum value.
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 5 2 ( 2 0 2 4 ) 1 1 1 3 e1 1 2 3 1121

Table 8 e Summary of the economic analysis for the optimum conditions: 18 MW of plant capacity, 190  C of reactor
temperature and 6 MPa of pressure (currency in Euro).
3,5-dibenzyl toluene 2-methyl indole
2 tanks parallel 2 tanks series 1 tank 2 tanks series
Reactor size (m3, each tank) 14 40 25 25
TOTAL CAPITAL INVESTMENT 3 700 000 6550 000 3 400 000 5070 000
ANNUALIZED INVESTMENT 740 000 1300 000 670 000 1 010 000
OPERATING COSTS 12 540 000 12 950 000 12 710 000 12 550 000
TOTAL ANNUAL COST 13 280 000 14 250 000 13 380 000 13560 000
H2 plant power (MW) 18.5 18.6 17.8 18.1
H2 consumption (t H2/year) 4860 4890 4680 4760
H2 storage cost (V/kg H2) 2.73 2.92 2.86 2.86

Process scale-up and economic analysis at optimum
conditions
The plant has been scale-up at the optimum conditions
determined in the sensitivity analysis of the previous section:
plant capacity of 18 MW, reactor temperature of 190  C and
pressure of 6 MPa. Table 8 summarizes the main economic
indicators and the H2 storage cost for the two LOHC that
performed better: 3,5-dibenzyl toluene and 2-methyl indole.
The reactor configurations with one and two tanks were
considered. However, for 3,5-dibenzyl toluene, the use of just
one single tank would require large tank size, exceeding the
maximum recommended. Therefore, the use of two tanks in
parallel of equal size was considered for this case.
The economic analysis for the optimum conditions in-
dicates that the process with the lowest capital investment is
the one based on 2-methyl indole with a reactor of one tank.
The operating costs are quite similar for all the processes, the
lowest one being that based on 3,5-dibenzyl toluene with a
reactor of two tanks in parallel. This process is also the one
with the lowest total annual cost and hydrogen storage cost,
2.73 V/kg. Consequently, this is the recommended process
and operating conditions of all the processes studied in this
work. The use of a reactor made of two tanks in parallel reduce
considerably the capital investment and makes it easier the
control of the reactor temperature.
The generation of hydrogen using renewable energy (e.g.,
solar and wind power) may suffer from intermittent
production with periods of surplus (for example, in spring-
summer) and periods of depletion, when the plant must be
stopped. The economic analysis has been carry out consid-
ering a 10-year period (with interest rate of 15%) to annualise
the capital investment and that the plant operates for 300 days
every year. In case the plant operated for fewer days a year,
the consumption of H2 and generation of LOHC would
decrease and, hence, the cost of H2 storage would increase. To
evaluate the impact of this, one should examine the contri-
bution of the different costs to the total annual costs. Thus,
94% of the total annual costs correspond to operating costs
(Table 8), of which 74% are due to raw materials (mainly LOHC
make-up) and utilities (Table 6). When the plant stops, raw
materials and utilities are not consumed. Since they are the
greater contributors to the total annual costs (70% of them)
and the H2 storage cost, the impact of fewer operating days a
year would be limited.
To verify this, the influence on the H2 storage cost of the
annual usage of the plant capacity has been evaluated for the
case of the 3,5-dibenzyl toluene process with a reactor of two
tanks in parallel (the process with the lowest H2 storage cost).
The nominal plant capacity is 18 MW, (Table 8). As shown in
Fig. 4a, if the annual use of plant capacity decreases to a 50% of
the nominal, the storage cost will increase ca. 20%.
In the calculations, it was assumed a LOHC recovery of 90%
considering all the stages, i.e., hydrogenation, storage, and
dehydrogenation. The LOHC recovery is negatively affected by
undesired reactions, taking place during the hydrogenation

Fig. 4 e Analysis of the influence of the annual use of plant capacity (a) and LOHC recovery yield (b) on the hydrogen storage
cost. Nominal plant capacity: 18 MW. LOHC: 3,5-dibenzyl toluene with a reactor of two tanks in parallel. Bar diagrams:
annualized investment and operating costs.
1122 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 5 2 ( 2 0 2 4 ) 1 1 1 3 e1 1 2 3
and dehydrogenation, and transforming the LOHC into other
compounds not suitable for hydrogen storage. One of the most
important criteria when selecting a LOHC candidate is their
high selectivity, to reduce the formation of these undesired
compounds.
As shown in the sensitivity analysis of Fig. 4b, the LOHC
recovery has an important impact on the final cost of
hydrogen storage. The operating costs, of which the most
important is the cost of the LOHC make-up, have a great
impact in the storage costs. Thus, for 3,5-dibenzyl toluene, an
increase of LOHC recovery from 90% to 99% would decrease
the storage cost from 2.73 to 1.50 V/kg (these figures corre-
spond to the optimized process). This increase in recovery is
plausible as the technology is completely developed and
stablished to industrial scale.
Conclusions
The adequacy of 3,5-dibenzyl toluene, 2-methyl indole and N-
ethylcarbazole as liquid organic hydrogen carriers (LOHC) has
been evaluated. The hydrogenation of these compounds in
three-phase stirred tank reactors, loaded with a slurry of Ni and
Ru-based catalysts, has been scaled-up. The proposed process
is formed by the following stages: compression, reaction, sep-
aration, and hydrogen recycle. An economic evaluation per-
formed on the process suggested that the most important costs
are the operating costs associated to raw materials (the LOHC)
and the investment cost corresponding to the reactor.
The influence of the main operating variables on the cost of
hydrogen storage has been evaluated by means of a sensitivity
analysis. The optimum variables are plant capacity, 18 MW,
reactor pressure, 6 MPa, and reactor temperature, 190  C. The
LOHC with the minimum hydrogen storage cost at optimum
operating conditions is 3,5-dibenzyl toluene with 2.73 V/kg.
Specifically, a reactor configuration made of two stirred tanks
in parallel was found to be optimum. It can be concluded that
the storage cost can be reduced considerably (even down to 1.5
V/kg) by increasing the LOHC recovery of the overall storage
process (made of hydrogenation, storage, and dehydrogena-
tion stages). Given the high selectivity of the involved re-
actions, this is plausible in the short term as the technology is
established at industrial scale.
Declaration of competing interest
The authors declare that they have no known competing
financial interests or personal relationships that could have
appeared to influence the work reported in this paper.
Acknowledgements
This work has been funded by the Spanish Ministry of Science
and Innovation, Spanish Agency for Research AEI (PID2020-
112587RB-I00) and the Regional Government of Asturias

n Hidro
(Misio
geno- AYUD/2021/9861).
references
[1] https://www.europarl.europa.eu/factsheets/en/sheet/72/
combating-climate-change.


[2] Alvarez M, Marı́n P, Ordo
n~ ez S. Production of dimethyl
carbonate by gas-phase oxidative carbonylation of methanol
over Cu/Y zeolite: mechanism and kinetics. Fuel Process
Technol 2023;247:107805.

ez R, Marı́n P, Ordo
[3] Pela
n
~ ez S. Direct synthesis of dimethyl
ether from syngas over mechanical mixtures of CuO/ZnO/
Al2O3 and g-Al2O3: process optimization and kinetic
modelling. Fuel Process Technol 2017;168:40e9.
€ ckl F, Schill W-P, Zerrahn A. Optimal supply chains and
[4] Sto
power sector benefits of green hydrogen. Sci Rep
2021;11:14191.
[5] Yee Mah AX, Ho WS, Hassim MH, Hashim H, Liew PY,
Muis ZA. Targeting and scheduling of standalone renewable
energy system with liquid organic hydrogen carrier as energy
storage. Energy 2021;218:119475.
[6] Liu W, Zuo H, Wang J, Xue Q, Ren B, Yang F. The production
and application of hydrogen in steel industry. Int J Hydrogen
Energy 2021;46:10548e69.
[7] Kojima Y, Yamaguchi M. Ammonia as a hydrogen energy
carrier. Int J Hydrogen Energy 2022;47:22832e9.
[8] Marı́n P, Patin~ o Y, Dı́ez FV, Ordo
n~ ez S. Modelling of hydrogen
perm-selective membrane reactors for catalytic methane
steam reforming. Int J Hydrogen Energy 2012;37:18433e45.
[9] Ma K, Deng J, Wang G, Zhou Q, Xu J. Utilization and impacts
of hydrogen in the ironmaking processes: a review from lab-
scale basics to industrial practices. Int J Hydrogen Energy
2021;46:26646e64.
[10] Ren J, Musyoka NM, Langmi HW, Mathe M, Liao S. Current
research trends and perspectives on materials-based
hydrogen storage solutions: a critical review. Int J Hydrogen
Energy 2017;42:289e311.
[11] Schlapbach L, Züttel A. Hydrogen-storage materials for
mobile applications. Nature 2001;414:353e8.
[12] Preuster P, Alekseev A, Wasserscheid P. Hydrogen storage
technologies for future energy systems. Annu Rev Chem
Biomol Eng 2017;8:445e71.
[13] Barthelemy H, Weber M, Barbier F. Hydrogen storage: recent
improvements and industrial perspectives. Int J Hydrogen
Energy 2017;42:7254e62.
[14] Ramirez-Vidal P, Canevesi RLS, Sdanghi G, Schaefer S,
Maranzana G, Celzard A, et al. A step forward in understanding
the hydrogen adsorption and compression on activated
carbons. ACS Appl Mater Interfaces 2021;13:12562e74.
[15] Fomkin A, Pribylov A, Men’shchikov I, Shkolin A, Aksyutin O,
Ishkov A, et al. Adsorption-based hydrogen storage in
activated carbons and model carbon structures. Reactions
2021;2:209e26.
[16] Thomas KM. Hydrogen adsorption and storage on porous
materials. Catal Today 2007;120:389e98.
[17] He T, Wu H, Wu G, Wang J, Zhou W, Xiong Z, et al.
Borohydride hydrazinates: high hydrogen content materials
for hydrogen storage. Energy Environ Sci 2012;5:5686e9.
[18] Juangsa FB, Irhamna AR, Aziz M. Production of ammonia as
potential hydrogen carrier: review on thermochemical and
electrochemical processes. Int J Hydrogen Energy
2021;46:14455e77.
[19] Heris M-N, Mirzaei MA, Asadi S, Mohammadi-Ivatloo B,
Zare K, Jebelli H, et al. Evaluation of hydrogen storage
technology in risk-constrained stochastic scheduling of multi-
carrier energy systems considering power, gas and heating
network constraints. Int J Hydrogen Energy 2020;45:30129e41.
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 5 2 ( 2 0 2 4 ) 1 1 1 3 e1 1 2 3
[20] Makepeace JW, He T, Weidenthaler C, Jensen TR, Chang F,
Vegge T, et al. Reversible ammonia-based and liquid organic
hydrogen carriers for high-density hydrogen storage: recent
progress. Int J Hydrogen Energy 2019;44:7746e67.
[21] Ferreira APRA, Oliveira RCP, Mateus MM, Santos DMF. A
review of the use of electrolytic cells for energy and
environmental applications. Energies 2023;16:1593.
[22] Bollmann J, Schmidt N, Beck D, Preuster P, Zigan L,
Wasserscheid P, et al. A path to a dynamic hydrogen storage
system using a liquid organic hydrogen carrier (LOHC):
burner-based direct heating of the dehydrogenation unit. Int
J Hydrogen Energy 2023;48:1011e23.
[23] Kim C, Lee Y, Kim K, Lee U. Implementation of formic acid as
a liquid organic hydrogen carrier (LOHC): techno-economic
analysis and life cycle assessment of formic acid produced
via CO2 utilization. Catalysts 2022;12:1113.
[24] Hamayun MH, Hussain M, Maafa IM, Aslam R. Integration of
hydrogenation and dehydrogenation system for hydrogen
storage and electricity generation e simulation study. Int J
Hydrogen Energy 2019;44:20213e22.
[25] Preuster P, Papp C, Wasserscheid P. Liquid organic hydrogen
carriers (LOHCs): toward a hydrogen-free hydrogen
economy. Acc Chem Res 2017;50:74e85.
[26] Niermann M, Beckendorff A, Kaltschmitt M, Bonhoff K.
Liquid organic hydrogen carrier (LOHC) e assessment based
on chemical and economic properties. Int J Hydrogen Energy
2019;44:6631e54.
[27] Lee S, Kim T, Han G, Kang S, Yoo Y-S, Jeon S-Y, et al.
Comparative energetic studies on liquid organic hydrogen
carrier: a net energy analysis. Renew Sustain Energy Rev
2021;150:111447.
[28] Markiewicz M, Zhang Y-Q, Empl MT, Lykaki M, Tho € ming J,
Steinberg P, et al. Hazard assessment of quinaldine-,
alkylcarbazole-, benzene- and toluene-based liquid organic
hydrogen carrier (LOHCs) systems. Energy Environ Sci
2019;12:366e83.
[29] Dı́az E, Rapado-Gallego P, Ordo
n~ ez S. Systematic evaluation
of physicochemical properties for the selection of alternative
liquid organic hydrogen carriers. J Energy Storage
2023;59:106511.
[30] Hiyoshi N, Rode CV, Sato O, Shirai M. Biphenyl
hydrogenation over supported transition metal catalysts
under supercritical carbon dioxide solvent. Appl Catal, A
2005;288:43e7.
[31] Leinweber A, Müller K. Hydrogenation of the liquid organic
hydrogen carrier compound monobenzyl toluene: reaction
pathway and kinetic effects. Energy Technol 2018;6:513e20.
[32] Huynh N-D, Hur SH, Kang SG. Tuning the dehydrogenation
performance of dibenzyl toluene as liquid organic hydrogen
carriers. Int J Hydrogen Energy 2021;46:34788e96.
[33] Hu P, Fogler E, Diskin-Posner Y, Iron MA, Milstein D. A novel
liquid organic hydrogen carrier system based on catalytic
peptide formation and hydrogenation. Nat Commun
2015;6:6859.

B, Byun M, Lim H. Design, economic evaluation,
[34] Brigljevic
and market uncertainty analysis of LOHC-based, CO2 free,
hydrogen delivery systems. Appl Energy 2020;274:115314.
1123
[35] Oh J, Jeong K, Kim TW, Kwon H, Han JW, Park JH, et al. 2-(N-
Methylbenzyl)pyridine: a potential liquid organic hydrogen
carrier with fast H2 release and stable activity in consecutive
cycles. ChemSusChem 2018;11:661e5.
[36] Yang X, Wu Y, Yu H, Sun M, Zheng J, Li X, et al. A YH3
promoted palladium catalyst for reversible hydrogen
storage of N-ethylcarbazole. Int J Hydrogen Energy
2020;45:33657e62.
[37] Fei S, Han B, Li L, Mei P, Zhu T, Yang M, et al. A study on the
catalytic hydrogenation of N-ethylcarbazole on the
mesoporous Pd/MoO3 catalyst. Int J Hydrogen Energy
2017;42:25942e50.
[38] Yang M, Dong Y, Fei S, Ke H, Cheng H. A comparative study of
catalytic dehydrogenation of perhydro-N-ethylcarbazole
over noble metal catalysts. Int J Hydrogen Energy
2014;39:18976e83.
[39] Gianotti E, Taillades-Jacquin M, Rozie re J, Jones DJ. High-
purity hydrogen generation via dehydrogenation of organic
carriers: a review on the catalytic process. ACS Catal
2018;8:4660e80.
[40] Brückner N, Obesser K, Bo € smann A, Teichmann D, Arlt W,
Dungs J, et al. Evaluation of industrially applied heat-transfer
fluids as liquid organic hydrogen carrier systems.
ChemSusChem 2014;7:229e35.
[41] Ali A, Kumar GU, Lee HJ. Investigation of hydrogenation of
Dibenzyltoluene as liquid organic hydrogen carrier. Mater
Today Proc 2021;45:1123e7.
[42] Shi L, Qi S, Qu J, Che T, Yi C, Yang B. Integration of
hydrogenation and dehydrogenation based on
dibenzyltoluene as liquid organic hydrogen energy carrier.
Int J Hydrogen Energy 2019;44:5345e54.
[43] Wan C, An Y, Xu G, Kong W. Study of catalytic hydrogenation
of N-ethylcarbazole over ruthenium catalyst. Int J Hydrogen
Energy 2012;37:13092e6.
[44] Li L, Yang M, Dong Y, Mei P, Cheng H. Hydrogen storage and
release from a new promising liquid organic hydrogen
storage carrier (LOHC): 2-methylindole. Int J Hydrogen
Energy 2016;41:16129e34.
[45] Sotoodeh F, Zhao L, Smith KJ. Kinetics of H2 recovery from
dodecahydro-N-ethylcarbazole over a supported Pd catalyst.
Appl Catal, A 2009;362:155e62.
[46] Runge P, So € lch C, Albert J, Wasserscheid P, Zo € ttl G,
Grimm V. Economic comparison of different electric fuels
for energy scenarios in 2035. Appl Energy 2019;233e234:
1078e93.
[47] Woods DR. Rules of thumb in engineering practice. Wiley-
VCH; 2007.
[48] Azzopardi BJ, Mudde RF, Lo S, Morvan H, Yan Y, Zhao D.
Sparged stirred vessels. Hydrodynamics of gas-liquid
reactors. 2011. p. 61e89.
[49] Biardi G, Baldi G. Three-phase catalytic reactors. Catal Today
1999;52:223e34.
[50] Towler G, Sinnott R. Chemical engineering design. Principles,
practice and economics of plant and process design. 2nd ed.
Elsevier; 2008.
[51] Peters MS, Timmerhaus KD, West RE. Plant design and
economics for chemical engineers. 5º ed. McGraw-Hill; 2003.